New Immune Dataset Targets Predictive Drug Discovery
Parse Biosciences and Nomic Bio launch a 100-million-cell database to accelerate the development of virtual immune models.
Updated on Oct. 29, 2026 in Biotech
Live Poll
Do you believe using AI-driven datasets will meaningfully accelerate the discovery of better medicines?
Parse Biosciences and Nomic Bio have announced a partnership to build the Cascade dataset, a collection of 100 million single-cell RNA-seq profiles paired with proteomic secretome data. This industry-scale dataset is designed to provide comprehensive mappings of human immune responses to various drug and cytokine conditions.
Why it matters
The project aims to overcome a significant bottleneck in drug discovery by providing the high-dimensional data needed to build accurate predictive virtual cell models. By linking gene expression to secreted protein output, the effort seeks to improve how researchers simulate how human immune cells react to potential therapies.
The Cascade dataset integrates 100 million single-cell RNA-seq profiles with over 11,000 proteomic secretome profiles. Each sample utilizes the Omini 1000 platform to measure up to 1,000 secreted proteins per sample across 150,000 combinatorial perturbation conditions.
The players
Parse Biosciences
A developer of single-cell sequencing technologies that provide scalable analysis of gene expression in complex tissues.
Nomic Bio
A proteomic technology firm known for its high-throughput Omini platform that measures protein signatures in biological samples.
GSK, Pfizer, Sanofi, Graph Therapeutics, and Turbine
Foundational pharmaceutical and biotechnology partners committed to utilizing the Cascade dataset for AI-driven drug discovery efforts.
The details
Parse Biosciences generates single-cell RNA-seq data—a method for measuring gene expression at the individual cell level—within their GigaLab facility. Nomic Bio contributes its Omini 1000 platform, a high-throughput proteomic technology that identifies and quantifies the proteins released by cells into their environment, known as the secretome. Together, these technologies correlate cellular instruction sets with functional protein output to map how immune systems shift under drug influence.
Timeline
October 29, 2026: Parse Biosciences and Nomic Bio announced the partnership.
The Tech Race
This initiative represents a strategic pivot in the race to digitize human biology for machine learning applications. It moves beyond the static mapping seen in the Human Cell Atlas to create a dynamic, perturbation-based training set for predictive pharmaceutical AI.
Pharmaceutical researchers at participating firms can access the dataset immediately for training predictive models. For the broader industry, the shift suggests that computational drug discovery will rely increasingly on massive, multi-omic datasets that capture cellular response in varied conditions.
The takeaway
The success of this initiative will be defined by whether it can reliably increase the speed and accuracy of candidate drug selection in internal pipelines. Watch for performance metrics from foundational partners like GSK and Sanofi in upcoming industry conferences to validate the utility of this dataset.
Further reading
Explore more developments in the Biotech sector regarding data-driven drug discovery.
More information
For details on partnership access, visit the licensing inquiries for Cascade dataset.
Live Poll
Do you believe using AI-driven datasets will meaningfully accelerate the discovery of better medicines?
